33% of Swedes feel unsafe about responsible AI use in healthcare

A YouGov survey commissioned by Tieto found that 33% of Swedes feel unsafe about whether artificial intelligence is used responsibly in Swedish healthcare. The survey included 1,022 people. Tieto Caretech’s Per Mattsson cautioned against drawing conclusions about what is behind the concern: the survey did not investigate its causes.

Views differ on how much autonomy AI should have. In the survey, 38% accepted AI analyzing X-rays and test results without human involvement, while 26% accepted it answering patient questions on its own. Only 19% said they could accept AI making diagnoses without healthcare staff involved.

Mattsson said AI can analyze large amounts of information and identify patterns to support decisions, while medical expertise and human interaction remain central. He sees potential for AI to summarize and structure data, reducing administrative work and freeing staff time for patients. He said the available data may be incomplete or contradictory, and needs systematic quality checks. AI also needs to fit existing healthcare workflows without compromising security.

On healthcare capacity, 48% thought AI could help address problems over the next five years. Of those, 41% viewed it as a complement to healthcare staff and 6% as a necessary part of the solution. Meanwhile, 43% said AI could not help: 23% considered it too immature, and 20% said human assessments are always needed. The survey was based on 1,022 online interviews with people over 18 in Sweden, conducted from 18 to 23 June 2026.